As a graduate student, I had the good fortune of meeting Dr Don Coffey, a professor of Urology, Oncology, Pharmacology, and Molecular Sciences at the Johns Hopkins School of Medicine, one of the most creative and humble scientists I have known.
I recall him telling a lunchtime gathering of us wide-eyed trainees that American graduate students don’t think about their experiments enough; that is, one should spend some 10-20% of the time it took to do the experiment thinking about the results, especially if the results are not what you expected.
The exercise he challenged us to do for “failed” experiments is to ask, “If this [result] is true, what does it imply?”
So, I thought of the good Dr Coffey last week when a highly-discussed paper appeared in JAMA describing a meta-analysis of antioxidant use in primary and secondary disease prevention trials. The surprising outcome was that, for some antioxidants like beta-carotene, vitamin A, and vitamin E, the relative risk of mortality increased between 4 and 16%. (abstract here but full article requires subscription). Vitamin C or selenium were associated with no change in mortality.
But very few pixels have been expended on how antioxidants might be increasing mortality if indeed the findings are true.
The very short two-fold answer is 1) that some antioxidants can be converted into more active pro-oxidants when sopping up free radicals and 2) free radicals are not always bad and, in fact, play important physiological defense roles in fighting infections and cancer.
Calling a compound an antioxidant is tricky business because one has to consider what happens to the so-called antioxidant in the process of making another molecule less reactive. Vitamin E is a perfect example of an antioxidant that can be turned into a pro-oxidant. If the pro-oxidant form of vitamin E is not reduced by another antioxidant, it can be damaging.
In terms of fighting infection, granulocytes normally produce an oxidative burst to help kill invading bacteria. Could interfering with this process with antioxidant vitamins worsen the progression of life-threatening infections?
Moreover, reactive oxygen species also mediate the mitochondrial pathway of apoptosis, or programmed cell death. While we often associate this process with how anticancer drugs work, normal apoptosis is also the reason that we don’t have a greater incidence of cancer – cells that have accumulated so much DNA damage that cannot be repaired (from sunlight, dietary carcinogens, etc.) are triggered to undergo apoptosis so that they don’t become cancer cells.
Mind you, these are each oversimplified explanations of steps where inappropriate action of an antioxidant could contribute to mortality and, for the most part, are theoretical.
So, while much more debate will continue on the drawbacks of the study design and the responses by supplement stakeholders (see this great essay by David Michaels for a discussion), it does pay to think for a moment or two what the JAMA study might be telling us about antioxidant supplementation and the perceived promise of improved health.